Variable

⧫ An item of data
⧫ Examples:
– gender
– test scores
– weight
⧫ Value varies from one observation to
another
Types/Classifications of Variables
⧫ Qualitative
⧫ Quantitative
– Discrete
– Continuous
 Qualitative Data
⧫ Describes the quality
⧫ Non-numerical format
Counts
Cannot order or measure
⧫ Examples
– gender
– marital status
– geographical region
– job title….
 Categorical data
⧫ Non-overlapping categories or
characteristics
⧫ Examples:
– Completes/Incompletes
– Professions
– Gender
 Quantitative Data
⧫ Frequencies
⧫ Measurements
 Discrete
⧫ Measurements are integers

⧫ Examples:
– number of employees of a company
– number of incorrect answers on a test
– number of participants in a program…
 Continuous
⧫ Measurements can take on any value -
usually within some range
⧫ Examples:
– Age
– Income
⧫ Arithmetic operations such as differences
and averages make sense.
Qualitatiave or Quantitative?
Discrete or Continuous?
⧫ Score on a placement exam
⧫ Preferred restaurant
⧫ Dollar amount of a loan
⧫ Height
⧫ Salary
⧫ Length of time to complete a task
⧫ Number of applicants
⧫ Ethnic origin
 Treatment as Ranks
⧫ Natural order
⧫ Not strictly measured
⧫ Examples:
– Age group
– Likert Scale data
⧫ Distinction between adjacent points on the
scale is not necessarily the same
 Analysis
Qualitative Data
⧫ Frequency tables
⧫ Modes - most frequently occurring
⧫ Graphs: Bar Charts and Pie Charts
 Analysis
Quantitative Data
⧫ Any form
⧫ Create groups or categories and generate
frequency tables
⧫ All descriptive statistics
 Effective Graphs:
Quantitative Data
⧫ Histograms
⧫ Stem-and-Leaf plots
⧫ Dot Plots
⧫ Box plots
⧫ XY Scatter Plots (2 variables).
Examples of Graphs
 Pie Chart

P e rfo rm a n c e A p p ra is a ls
10%

38%
M o re D iffic u lt
14%
D iffic u lt
S am e
M u c h E a s ie r
E a s ie r

33%
90
80
70
60
50 E a st
40 W e st
30 N o r th
20
10
0
1 st Q tr 2 n d Q tr 3 r d Q tr 4 th Q tr
 Histogram
H isto g ram

12
10
8
Frequency

6
4
2
0
49 59 69 79 89 99
Sco r e
 Boxplot
B o x p lo t o f C 1

20 30 40 50 60 70 80 90 100 110

C1
Stem and Leaf Plot
StemandLeaf Plot
W eight of Meat

75
83
87999
923
966789
10
10688
112244
11788
124
128
13
138
141
 Analyze Ranked Data
⧫ Frequency tables
⧫ Mode, Median, Quartiles
⧫ Graphs:
– Bar Charts
– Dot Plots, Pie Charts
– Line Charts (2 variables)
 Data Example
Suggest some ways you could analyze these items.

⧫ Score on a placement exam

⧫ Preferred restaurant
⧫ Dollar amount of a loan
⧫ Height
⧫ Salary
⧫ Length of time to complete a task
⧫ Number of applicants
⧫ Ethnic origin
 Tables and Graphs
Note Excel will create any graph that you
specify

Consider the type of data before selecting

your graph.
 Frequency Table/Frequency
Distribution
Summarize data:
⧫ categorical
⧫ nominal
⧫ Continuous data - the data set has been
divided into meaningful groups
 Frequency Distribution
Count the number of observations that fall
into each category.

Frequency: the number associated with each

category
 Relative Frequency Distribution
Proportion of observations falling in a given
category

Report relative frequencies or percentages

 Example
Frequency Distribution
No. of Defective Parts n
0 &< 2 0
2 4 4
4 6 5
6 8 12
8 10 14
10 12 9
12 14 10
14 16 6
 Graphs
Categorical/Qualitative Data
 Pie Charts
⧫ Circle - divided Viet Swedish
proportionately Namese 5%
15%

English
Spanish 55%
⧫ Segment - percentage of 25%
the whole that falls into
each category Native Language
 Bar Charts
⧫ Bar charts - % in various
categories A
v
er
age
Uni
tsS
ol
d (
per
per
son
)b
yP
ro
du
ct

⧫ Vertical scale - 2
0

Average Sold/Person
frequencies, relative 1
5

frequencies 1
0

⧫ Horizontal scale -
5 B
ef
o r
e T
r
aini
n g

A
ft
erT
r
aini
n g
0
categories B
41 B
A4
2 B
41
F C
21 O
t
her

⧫ Allows comparisons
P
r
odu
ct
 Constructing Bar Charts
⧫ All boxes should have the same width
⧫ Gaps between the boxes - no connection
between
⧫ Any order.
⧫ Use to represent two categorical variables
simultaneously
 Graphs: Measured
Continues Quantitative Data
⧫ Histograms
⧫ Stem and Leaf
⧫ Box plots
⧫ Line Graphs
⧫ XY Scatter Charts (2 variables)
 Histograms
⧫ Frequency G
r
a
d
eD
is
t
r
i
bu
t
io
n
distributions of 1
2
1
0
continuous variables

Frequency
8
6

⧫ Drawn without gaps 4

2
0 G
r
ad
e
between the bars 5
96
97
98
99
9
 Constructing Histograms
⧫ Non-overlapping
intervals G
r
ade
Di
str
ib
ut
ion

⧫ Intervals - generally 1
2
1
0
the same length

Frequency
8
6
⧫ Number of values in 4
2
each interval -class 0 G
r
ade
5
9 6
9 7
9 8
9 9
9
frequency
⧫ Relative frequencies o
 XY Scatter Chart
⧫ Two variables
Abscent by Age
⧫ Variables: quantitative and
20
continuous.

Days Absent
15
⧫ Plot pairs - rectangular
10
coordinate system 5
⧫ Examine the relationship 0
0 10 20 30 40 50 60 70
between two variables
Age
 Line Chart
⧫ Similar to the scatter
1997MonthlySales
chart
⧫ Values of the 170
165

Sales (x$10,000)
160
independent variable 155
150
(shown on the 145
140
horizontal axis) can be 135
130
ranked values (i.e.. 125

Feb

June
Jan

Mar

May
Apr
they do not have to be Month

continuous variables).
Basic Principles for Constructing
All Plots
⧫ Data should stand out clearly from
background
⧫ The information should be clearly labeled
– title
– axes, bars, pie segments, etc. - include units that
are needed to interpret data
– scale including starting points.
 Principles cont.
⧫ Source
⧫ No clutter
⧫ Minimize information or data on one graph.
⧫ Try several approaches
 Describing Data
⧫ Shape of the Distribution
– Symmetry
– Skewness
– Modality: most frequently occurring value
– Unimodal or bimodal or uniform
 e

e
Right Skewed
r
Left Skewed

r
H i s t o g r a m H i s t o g r a m

1 2 1 2
1 0 1 0
8 8
6 6
4 4
F

F
2 2
0 0
5 9 6 9 7 9 8 9 9 9 5 9 6 9 7 9 8 9 9 9

G r a d e G r a d e

H
is
tog
ram

1
2
1
0
Frequency

8
6
4 Symmetrical
2
0
5
9 6
9 7
9 8
9 9
9
G
rad
e
 Describing Data
⧫ Centrality
⧫ Spread
⧫ Extreme values
 Measures of Centrality
⧫ Mean
⧫ Median
⧫ Mode
 Mean
⧫ Most common measure
⧫ Extremely large values in a data set will
increase the value of the mean
⧫ Extremely low values will decrease it.
 Calculating the
Mean
T1 T2 T3
85 85 85
90 90 90
75 35 75
90 90 110
340 300 360 Sum
85 75 90 Mean
 Median
⧫ Central point .
⧫ Half of the data has a value than the median
⧫ Half of the data has a higher value than the
median
⧫ Not affected by extremely large or small
values
 Find the Median
85 90 75 92 95 Data

75 85 90 92 95 Sorted Data

Median is 90.
 Find the Median
95 90 92 85 Data

85 90 92 95 Sorted Data

Median:
(90 + 92)/2 = 91
Measures of Spread
 Range
⧫ Subtract the smallest value from the largest
⧫ Report the smallest and largest values.

85 90 75 92 95 Scores
Range: 75 to 95
or 20
 Variance/Standard Deviation
⧫ Average variation of the data values from
the mean of the values
⧫ Variance.
 The Empirical Rule
⧫ Symmetrical Data
⧫ At least:
68% of the data values are within one standard
deviation of the mean
90% of the data values are within two standard
deviation of the mean
99% of the data values are within three standard
deviations of the mean
 Tchybychef’s Inequality
⧫ Skewed Data
⧫ At least:
75% of the data values are within two standard
deviation of the mean.
90% of the data values are within one standard
deviation of the mean.
 Measures of Relative Standing
⧫ Percentiles

⧫ Quartiles
 Quartiles
⧫ The lower quartile is the same as the 25th
percentile.
– 25% of the scores are lower and
– 75% of the scores are higher than the lower
quartile.
⧫ The upper quartile is the same as the 75th
percentile.
– 75% of the scores are lower and
 Correlation
Describes the strength of the relationship
between two (or more) variables

Pearson Product-moment Correlation

Coefficient - assumes continuous
quantitative data
 Relationship between Variables
⧫ Positive
⧫ Negative
⧫ No relationship.
 Interpreting Correlation
Coefficients.
⧫ 0.20 to 0.35- show a slight relationship
(little value in practical prediction situations)
⧫ 0.50 - crude group prediction
(Correlations this low do not suggest a good relationship)
⧫ 0.65 to 0.85 - group predictions that are good
⧫ Over 0.85 - a close relationship between the two
variables.
 Even a high correlation
coefficient does not establish a
cause and effect
relationship!!!!!
 Coefficient of Determination
⧫ Square root of the correlation coefficient

⧫ Gives the percent of variation in the

dependent variable that is ‘explained’ by the
independent variable.
⧫ Look at an XY scatter plot
 Least Square Line
⧫ Describe the relationship between the two
variables
⧫ Make predictions of the dependent variable
from the independent variable
 Positive Relationship
r will be a positive
8
7
number.
6
5
Y

4
3
2
1
0
0 2 4 6
X
 Negative Relationship

8
r will be a negative
7 number.
6
5
Y

4
3
2
1
0
0 2 4 6
X